saikat Gochhait
@siu.edu.in
Assistant Professor Academic Level 12 7th Pay CPC
Symbiosis International Deemed University
Dr. Saikat Gochhait teaches at Symbiosis Institute of Digital & Telecom Management, Symbiosis International Deemed University Pune, India and Neurosciences Research Institute-Samara State Medical University, Russia. He is Ph.D and Post-Doctoral Fellow from the UEx, Spain and National Dong Hwa University, Taiwan. He was Awarded DITA and MOFA Fellowship in 2017 and 2018. His research publication with foreign authors is indexed in Scopus, ABDC, and Web of Science. He is a Senior IEEE member.
EDUCATION
Post Doctoral Fellow - Uex, Spain
Post Doctoral Fellow - National Dong Hwa University, Taiwan
PhD - Sambalpur University
RESEARCH INTERESTS
Technology Management
Marketing
Healthcare
Entrepreneurship
FUTURE PROJECTS
Neurosciences
NeuroMarketing
Applications Invited
Collaborators
Entreprenuership
Women Entrepreneurs
Applications Invited
Collaborators
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Saikat Gochhait, Deepak K. Sharma, and Mrinal Bachute
University of Basrah - College of Engineering
Accurate long-term load forecasting (LTLF) is crucial for smart grid operations, but existing CNN-based methods face challenges in extracting essential features from electricity load data, resulting in diminished forecasting performance. To overcome this limitation, we propose a novel ensemble model that integrates a feature extraction module, densely connected residual block (DCRB), long short-term memory layer (LSTM), and ensemble thinking. The feature extraction module captures the randomness and trends in climate data, enhancing the accuracy of load data analysis. Leveraging the DCRB, our model demonstrates superior performance by extracting features from multi-scale input data, surpassing conventional CNN-based models. We evaluate our model using hourly load data from Odisha and day-wise data from Delhi, and the experimental results exhibit low root mean square error (RMSE) values of 0.952 and 0.864 for Odisha and Delhi, respectively. This research contributes to a comparative long-term electricity forecasting analysis, showcasing the efficiency of our proposed model in power system management. Moreover, the model holds the potential to sup-port decision making processes, making it a valuable tool for stakeholders in the electricity sector.
Shashank Mittal, Priyank Kumar Singh, Saikat Gochhait, and Shubham Kumar
IGI Global
Accurate disease prognosis is crucial for improved healthcare outcomes. Artificial intelligence (AI) offers immense potential in this domain, but traditional “black-box” models lack interpretability. This chapter explores the integration of Explainable AI (XAI) with Green AI, a resource-efficient and sustainable approach to AI development. They discuss how XAI can enhance trust in Green AI models for disease prognosis, mitigate potential biases, and promote responsible AI development. They highlight the challenges of balancing interpretability with efficiency and propose future research directions to unlock the full potential of XAI for Green AI-powered disease prognosis. This approach has the potential to revolutionize healthcare by providing accurate, transparent, and environmentally friendly tools for early disease detection and improved patient outcomes.
Mohit Yadav, Priyank Kumar Singh, Saikat Gochhait, Nisha Gaur, and Puwakpitiyage Gayan Dhanushka Wijethilaka
IGI Global
This chapter explores the potential of green AI and big data informatics for personalized disease prediction in clinical decision making. Green AI prioritizes efficiency, minimizing computational resources needed to analyze vast healthcare datasets. Big data informatics provides the platform to manage and analyze these datasets for knowledge discovery. This chapter delves into how green AI algorithms optimize resource utilization while big data platforms leverage diverse patient data for more accurate, individual risk assessments. The applications in clinical decision-making encompass early detection, risk stratification, and personalized treatment plans. However, ethical considerations regarding data privacy, bias, and potential job displacement require careful attention. Finally, the future directions highlight advancements in green AI efficiency, explainable models, and integration with other health technologies, paving the way for a future of proactive healthcare and patient empowerment.
Priyank Kumar Singh, Mohit Yadav, Saikat Gochhait, and P. G. S. Amila Jayarathne
IGI Global
In this chapter, the authors aim to discuss the significance of integrating AI prediction and green computing in the healthcare field to improve disease diagnosis, treatment, and patient care and minimise the adverse effects on the environment. The methodology employed is the systematic literature review (SLR) approach. The results show that combining green practices with AI prediction enhances the effectiveness and sustainability of the healthcare system. Practical implications are that there is a need for frequent policy updates and practical staff training to improve environmental management. The authors focus on the real-world implications and provide tactical recommendations for healthcare organisations that want to adopt green computing strategies successfully. A strategic perspective should be used with top management's support and all employees' involvement to achieve the organisation's future vision regarding these measures.
Shashank Mittal, Priyank Kumar Singh, Saikat Gochhait, Nisha Gaur, and Shubham Kumar
IGI Global
Clinical trial design is undergoing a revolution fueled by artificial intelligence (AI) and translational bioinformatics. This chapter explores how AI techniques like machine learning and deep learning are being harnessed to analyze vast datasets of biological and clinical information. By integrating these insights with translational bioinformatics, researchers can identify promising drug candidates, select patients most likely to benefit from treatment, and design more efficient and targeted clinical trials. Real-world examples showcase the application of AI in immuno-oncology patient selection, drug discovery for rare diseases, predicting Alzheimer's trial outcomes, and virtual patient recruitment for cardiovascular studies. While challenges like data quality and ethical considerations exist, AI and translational bioinformatics hold immense promise for accelerating drug development, bringing life-saving therapies to patients faster.
Shashank Mittal, Priyank Kumar Kumar Singh, Saikat Gochhait, and Shubham Kumar
IGI Global
AI is rapidly transforming the field of epidemiology. This chapter explores how AI integrates data analysis, predictive modeling, disease surveillance, and diagnostic tools to significantly improve public health outcomes. AI-driven methodologies enhance diagnostic accuracy, improve disease surveillance efficiency, and aid in developing better predictive models, all of which contribute to improved public health strategies. AI seamlessly integrates with traditional epidemiological approaches, paving the way for a new era in combating infectious diseases. Advancements in AI hold immense promise for the future of public health, with possibilities for real-time disease surveillance, personalized medicine, and more accurate predictive modeling. However, broader adoption and responsible use of AI require careful consideration of ethical issues, data privacy concerns, and collaboration among stakeholders. Ultimately, leveraging AI effectively has the potential to improve public health outcomes, ensure equitable access to healthcare, and enhance global preparedness for health crises.
Priyank Kumar Singh, Mohit Yadav, Saikat Gochhait, and Puwakpitiyage Gayan Dhanushka Wijethilaka
IGI Global
The burgeoning field of AI-powered healthcare prognosis offers immense potential, but traditional data center infrastructure creates a significant environmental footprint. This chapter advocates for energy-efficient AI algorithms and hardware alongside renewable energy integration (solar, wind) to minimize reliance on fossil fuels. Robust security measures and privacy-preserving techniques are crucial to protect sensitive patient data used in AI models. Finally, scalable cloud-based infrastructure with containerization and auto-scaling ensures efficient handling of growing data volumes and user demands. By prioritizing these principles, we can create a sustainable and secure future where AI empowers healthcare prognosis, improving patient outcomes for generations to come.
Saikat Gochhait
IGI Global
CDN is constituted of three basic components. A content provider is somebody entrusting the URI namespace of the Web objects to be dispersed. The content provider's server contains all such objects. A CDN provider can be some owner party that enables transportation conveniences to content providers to deliver content in a timely and reliable manner. They may employ geographically distributed caching and/or replica servers (surrogates or edge servers) to duplicate content. Together they may form what we call a web cluster. End users are the customers who use content from the content provider's website.
Saikat Gochhait
IGI Global
Cloud computing is gaining momentum as a subscription-oriented paradigm providing on-demand payable access to virtualized IT services and products across the net. It is a breakthrough technology that is offering on-demand access to various services across the network. Auto-scaling, though quite an attractive proposition to customers and naïve cloud service providers, has its own share of issues and challenges. This work was an attempt to classify and appreciate the auto scaling framework while outlining its challenges. Many effective and efficient auto scaling strategies are being deployed by cloud giants like Amazon AWS, Microsoft Azure, etc.
Saikat Gochhait, Yogesh Singh Rathore, Irina Leonova, Mahima Shanker Pandey, Bal Krishna Saraswat, Santosh Kumar Maurya, Hare Ram Singh, and Nidhi Bansal
Institute of Advanced Engineering and Science
<p>URL stands for uniform resource locator are the addresses of the unique resources on the internet. We all need URLs to access any type of resource on the internet, such as any web page, and document. Sometimes URLs can be long, irrelative and unattractive and unable to send sometimes via email. So, for this, we proposed a URL shortener web application based on the Python-Django platform which is fast and makes your long URLs in the shortest form which you can share on social media platforms. It makes all the messy, unattractive URLs short and shareable. Writing paper proposed a premium section in our application that gives access to the customizable URLs and analytics of your shorten URLs. Customizable URLs are the URLs you create by your own keywords. By creating a premium profile with the application, you can create your own URLs by using your own keywords. We have considered security a major part of the application that prevents the short URLs from being hacked or redirected to any advertising website or content. We store all the data related to the URL to show you the best view of your analytics and update it regularly. Main contribution in this field that for web application that provides users with a fast, secure and shortest URL for their using long URLs. Comparatively to other services that are currently available, the application provides superior security, availability, and confidentiality.</p>
Osama Al-Baik, Saleh Alomari, Omar Alssayed, Saikat Gochhait, Irina Leonova, Uma Dutta, Om Parkash Malik, Zeinab Montazeri, and Mohammad Dehghani
MDPI AG
A new bio-inspired metaheuristic algorithm named the Pufferfish Optimization Algorithm (POA), that imitates the natural behavior of pufferfish in nature, is introduced in this paper. The fundamental inspiration of POA is adapted from the defense mechanism of pufferfish against predators. In this defense mechanism, by filling its elastic stomach with water, the pufferfish becomes a spherical ball with pointed spines, and as a result, the hungry predator escapes from this threat. The POA theory is stated and then mathematically modeled in two phases: (i) exploration based on the simulation of a predator’s attack on a pufferfish and (ii) exploitation based on the simulation of a predator’s escape from spiny spherical pufferfish. The performance of POA is evaluated in handling the CEC 2017 test suite for problem dimensions equal to 10, 30, 50, and 100. The optimization results show that POA has achieved an effective solution with the appropriate ability in exploration, exploitation, and the balance between them during the search process. The quality of POA in the optimization process is compared with the performance of twelve well-known metaheuristic algorithms. The simulation results show that POA provides superior performance by achieving better results in most of the benchmark functions in order to solve the CEC 2017 test suite compared to competitor algorithms. Also, the effectiveness of POA to handle optimization tasks in real-world applications is evaluated on twenty-two constrained optimization problems from the CEC 2011 test suite and four engineering design problems. Simulation results show that POA provides effective performance in handling real-world applications by achieving better solutions compared to competitor algorithms.
Tushar Bharti, Shashwat Singh, Ved Prakash Chaubey, Shivangini Gupta, Shamneesh Sharma, and Saikat Gochhait
IEEE
In today's interconnected world, communication gaps still exist between individuals with sensory impairments, specifically the blind and deaf-mute communities. This paper introduces BridgeAI, a novel assistive technology platform designed to overcome these communication barriers by leveraging cutting-edge AI technologies. BridgeAI employs Convolutional Neural Networks (CNN) for sign language recognition, speech-to-text, and text-to-speech models tailored for communication between the blind and deaf-mute individuals. The key innovation lies in the use of a custom-built dataset, enhanced through MediaPipe for real-time hand tracking, alongside an integrated Region of Interest (ROI)-based approach to improve gesture recognition accuracy. This study presents the architecture, implementation, and results of BridgeAI, which achieved a sign language recognition accuracy of 96.8%, demonstrating significant potential in improving accessibility for the blind and deaf-mute communities.
Sanjeev Thakur, Rohit Bele, Tarun Yarlagadda, Ved Prakash Chaubey, Shamneesh Sharma, and Saikat Gochhait
IEEE
This paper analyzes the Speech Emotion Recognition (SER) procedure, a tool that helps IT systems increase their intelligence by recognizing persons' emotions from the signals put into words, by employing deep learning methods. The audio data was preprocessed by extracting the Mel-frequency cepstral coefficients (MFCCs) and later augmented by adding white noise, time shifting, pitch shifting and time stretching. The models used in this paper are Long Short-Term Memory (LSTM) networks, and Convolutional Neural Networks (CNNs). CNN architecture has convolutional operation layers, batch normalization layers followed by max-pooling layers, and dropout layers to learn speech features and to avoid overfitting. With these results, we stress the importance of the synergy of CNNs with solid preprocessing and augmentation methods in the context of SER and envision their tremendous potential in HCI, education, and health. It is worth stressing that including models such as SVM and LSTM certifies the novelty and the genetic soundness of the approach.
Shamik Tiwari, Saikat Gochhait, and Hithyshee Bonuga
IEEE
Gestational diabetes mellitus (GDM) is a con-dition in which a hormone produced by the placenta disrupts the body's ability to use insulin efficiently. Early diagnosis of GDM risk is essential since it can lead to more effective treatments and less cumulative effects. The purpose of this study aims to investigate an efficient model for early identification of GDM using commonly accessible parameters to enable early intervention. The project aims to use prediction algorithms in locations where more thorough assessments are unavailable. Multiple classifiers are created that function well in diagnosing GDM having accuracies ranging from 93% to 90.5%. KNN and XGB classifiers achieved 95% accuracy, which is the best among all classifiers. This suggests employing these classifiers to create a GDM detection system.
Sandipan Das, Subir Gupta, Asis Kumar Bhunia, Saikat Gochhait, Jayashri Deb Sinha, and Joyjit Patra
IEEE
This study presents a novel evaluation of the Part-Integrated Bedside Diagnostic Unit (PIBDU), designed for home healthcare use, emphasizing its dual focus on usability and effectiveness. The research addresses a critical gap in current healthcare technology by integrating both functional performance and user experience assessments. The findings demonstrate that the PIBDU is highly user-friendly, with 85% of participants finding it easy to use and 90% rating it as accessible. The study's innovation lies in its comprehensive methodology, which combines quantitative and qualitative data to provide a holistic evaluation of the device. This approach not only validates the PIBDU's reliability but also highlights the importance of user familiarity with technology, suggesting that enhanced training can further improve outcomes. These results underscore the device's potential to significantly impact home healthcare, offering valuable insights for future technological developments in this field.
Prabakaran Raghavendran, Tharmalingam Gunasekar, and Saikat Gochhait
IEEE
This paper utilizes an algorithm to encrypt and decrypt the message by using the Pourreza transformation integral transform and the congruence modulo operator. An example illustrates how such a mathematical tool would be applied in reality to provide secure message transmission through encrypting communications. This investigation highlights the strength of these changes and also makes the role of decision-making processes correlate with the integration of the Pourreza transformation with modulo operations leading to enhanced levels of confidentiality and integrity in data. Therefore, extended testing followed by an analysis puts this investigation of innovative cryptographic encoding and decoding designs before its ability to enhance security and knowledge in decision-making strategies while secret information is being transmitted.
Kartik Mathur, Nakul Mathur, and Saikat Gochhait
IEEE
This research focuses on binge-viewing among the young users related to OTT streaming services like Netflix and Amazon Prime. This study, having employed factor analysis, snowball sampling, social network, and A/B testing to determine the causes of binge-watching, reveals that individuals binge- watch because of; Variety, ease, escapism, and social interactions. Thus, analysis indicates the strong impact of such factors as modern technology and internet access on viewing pattern. The study does include some possible negative effects, which are associated with procrastination and health problems; it is again underlined that the content should be balanced. The findings of this research could help OTT platforms improve the content plan and apply better engagement by identifying the motives before the binge-sessions and gratifications after them. Hence, these research contributions can inform the enhancement of digital media delivery to respond to the youthful demographic's changing expectations and lessen the detrimental impact of streaming.
P. D. Vaidya, Saikat Gochhait, Tharmalingam Gunasekar, and Prabakaran Raghavendran
IEEE
Demand Side Management (DSM) is a critical strategy in the improvement of efficiency and reliability of the electrical system in smart grids. One of the effective DSM strategies is load shifting, which involves changing the timing of electricity consumption to off-peak hours. This paper gives a new approach to load shifting within residential areas through the Tunicate Swarm Algorithm. The TSA is an algorithm inspired by the swarming behavior of marine organisms called tunicates. These organisms are well known for their efficiency in movement and resource use. It is desired, within the scope of the proposed method, to optimize schedules for household appliances towards minimal electricity cost and minimal peak demand. TSA-based load-shifting technique: simulation result comparison with classical techniques—genetic algorithm and particle swarm optimization. It should be stated that the simulation results show a comparison of the performance of this proposed TSA-based technique about cost reduction and the mitigation of peak demand, which outperformed both GA and PSO. The results obtained in this study demonstrated that the TSA-based technique offers much potential for effective residential load management in smart grids, where quantifiable benefits could be derived in terms of energy savings and system stability. In the future, work will focus on real-world implementation and integration with other DSM strategies to enhance the overall efficiency of the electrical grid. This research adds to the ever-increasing bank of bio-inspired algorithms for energy management and acts as a springboard toward further areas of investigation in the field.
Shamik Tiwari, Saikat Gochhait, and Ritam Chatterjee
IEEE
Amyotrophic Lateral Sclerosis (ALS) is a persistently progressing neurological disease with restricted treatment choices. The advent of extensive global datasets and advanced machine learning models offers new opportunities to evaluate potential prognostic, inspection, and diagnostic indicators. Additionally, emerging categorization and staging systems aim to accurately stratify patients into distinct prognostic categories. This study evaluates several prominent machine learning classifiers for ALS classification. These experiments have exposed that the CatBoost classification algorithm attained the maximum performance, with an accuracy of 0.85 and an AUC of 0.97. Other significant models include XGB, Random Forest, and Extra-Trees classifiers, each showing an accuracy of around 0.75 but with varying AUC values.
Ngrddy Jhnvi, S.K. Bhrdwj, Ajay Sharma, Shamneesh Sharma, Ankur Sodhi, and Saikat Gochhait
IEEE
Human Activity Recognition for Fight Detection is an important research domain aimed at automatically identifying patterns indicative of physical altercations. Leveraging deep learning models like CNNs and RNNs, this approach extracts features from video frames to recognize fight-related behavioural patterns. The primary objective is to develop a machine learning model capable of autonomously detecting instances of fights in video footage, using techniques such as the Long-term Recurrent Convolutional Network (LRCN). Training involves a comprehensive dataset encompassing examples of both fights and non-fight activities, with model performance evaluated using standard metrics. Each frame undergoes individual analysis by the model to predict the presence of any indications of a fight. The model predicts the action with an accuracy of 98.03% and the movements given as input can be categorized as fights or no fights.
Syed Aman Hussain, Nareddy Yashwanth Reddy, Junnuthula Srivardhan, Ajay Sharma, Shamneesh Sharma, and Saikat Gochhait
IEEE
In the era of digital communication, it is more about being able to detect emotions than anything else if we are to maintain this complicated and demanding human society. Emotion recognition research is pulling out all the stops, resulting in all manner of methods to recognize human feelings. Emotions are often transmitted through words and body language, and facial expressions play an important part in communicating them. In this paper, the use of a Convolutional Neural Network (CNN) is proposed as a real-time emotion recognition system based on machine learning algorithm, and the system is able to process video feeds. The classifier breaks emotions down into seven categories, with an accuracy of about 81 percent. In addition, the research extends to the use of various image datasets to improve emotion forecasting accuracy.
RECENT SCHOLAR PUBLICATIONS
MOST CITED SCHOLAR PUBLICATIONS
GRANT DETAILS
Department of Science and Industrial Research , Govt of India with Grant of Rs 13,000,00
Ministry of Foreign Affairs, Taiwan with Grant of Rs 12,000,00
University of Deusto, Spain with Research Grant of Rs 2,000,00
University of Extremadura, Spain with Research Grant of Rs 2,000,00
Samara State Medical University, Russia with Research Visit grant of Rs 2,500,00
Symbiosis International Deemed University with Travel and Research Grant of 4,000,000
INDUSTRY EXPERIENCE
IFGL Refractories Ltd